Creation of new automated monitoring systems and diagnostics: detection of latent attacks on digital information and control systems of energy facilities

The article deals with the problems of ensuring the safety of information and control systems of power facilities. The necessity of developing a performance-scalable software and hardware complex for automated monitoring and diagnostics with a neural network component has been substantiated. The key characteristics of the complex, the tasks to be solved, the architecture, the main methods of data analysis to identify signs of unauthorized interference in the operation of functional units of power equipment are formulated. The effectiveness of neural networks in the tasks of identifying an attack, forming a strategy for blocking an attack, deciding on the introduction of a critical situation mode with the transition to technical means of non-programmable logic, as well as predicting the behavior of target functional units of equipment of energy facilities for the subsequent return to normal modes and restoration of equipment operability is shown.